Biometric personal identification system based on iris analysis

• US 5,291,560 A
• Filed: 07/15/1991
• Issued: 03/01/1994
• Est. Priority Date: 07/15/1991
• Status: Expired due to Term

• Current Assignee: IriScan, Inc. (Advent International Corporation)
• Original Assignee: IRI SCAN Incorporated
• Inventors: Daugman, John G.
• Primary Examiner(s): Mancuso, Joseph
• Application Number:
  US07/729,638
• Time in Patent Office: 960 Days
• Field of Search: 382/2, 382/6, 382/30, 382/9, 351/206, 351/221, 351/208, 351/205, 354/62, 364/413.01, 340/825.34
• US Class Current: 382/117
• CPC Class Codes: A61B 3/1216 for diagnostics of the iris
• CPC Class Codes: A61B 5/1171 based on the shapes or appearances of their bodies or parts thereof
• CPC Class Codes: G06F 21/32 using biometric data, e.g. fingerprints, iris scans or voiceprints
• CPC Class Codes: G06V 10/42 Global feature extraction by analysis of the whole pattern, e.g. using frequency domain transformations or autocorrelation
• CPC Class Codes: G06V 10/443 by matching or filtering
• CPC Class Codes: G06V 40/18 Eye characteristics, e.g. of the iris
• CPC Class Codes: G06V 40/45 Detection of the body part being alive
• CPC Class Codes: G07C 9/37 using biometric data, e.g. fingerprints, iris scans or voice recognition

Abstract

A system for rapid and automatic identification of persons, with very high reliability and confidence levels. The iris of the eye is used an optical fingerprint, having a highly detailed pattern that is unique for each individual and stable over many years. Image analysis algorithms find the iris in a live video image of a person'"'"'s face, and encode its texture into a compact signature, or "iris code." Iris texture is extracted from the image at multiple scales of analysis by a self-similar set of quadrature (2-D Gabor) bandpass filters defined in a dimensionless polar coordinate system. The sign of the projection of many different parts of the iris onto these multi-scale quadrature filters, determines each bit in an abstract (256-byte) iris code. The degrees-of-freedom in this code are based on the principle forms of variation in a population of irises studied. Because of the universal mathematical format and constant length of the iris codes, comparisons between them are readily implemented by the Exclusive-OR (XOR) logical operation. Pattern recognition is achieved by combining special signal processing methods with statistical decision theory, leading to a statistical test of independence based on a similarity metric (the Hamming distance) that is computed from the XOR of any two iris codes. This measure positively establishes, confirms, or disconfirms, the identity of any individual. It also generates an objective confidence level associated with any such identification decision.

Claims

1. A method for uniquely identifying a particular human being by biometric analysis of the iris of the eye, comprising the following steps:

acquiring an image of an eye of the human to be identified;

isolating and defining the iris of the eye within the image, wherein said isolating and defining step includes the steps of;

defining a circular pupillary boundary between the iris and pupil portions of the image;

defining another circular boundary between the iris and sclera portions of the image, using arcs that are not necessarily concentric with the pupillary boundary;

establishing a polar coordinate system on the isolated iris image, the origin of the coordinate system being the center of the circular pupillary boundary, wherein the radial coordinate is measured as a percentage of the distance between the said circular pupillary boundary and said circular boundary between the iris and sclera; and

defining a plurality of annular analysis bands within the iris image;

analyzing the iris to generate a presenting iris code;

comparing said presenting code with a previously generated reference iris code to generate a measure of similarity between said presenting iris code and said reference code;

converting said similarity measure into a decision that said iris codes either do or do not arise from the same iris; and

calculating a confidence level for the decision.

2. The method of claim 1, wherein:

said analysis bands exclude certain preselected portions of the iris image likely to be occluded by the eyelids, eyelashes, or specular reflection from an illuminator.

3. The method of claim 2, wherein said analyzing step includes the steps of:

analyzing the portion of the iris image lying within said annular analysis bands and employing signal processing means to generate an iris code for said iris image portion.

4. The method of claim 3, wherein:

said signal processing means comprises applying multi-scale, self-similar, two dimensional quadrature bandpass filters in polar coordinates to the iris image.

5. The method of claim 4, wherein:

said iris code has a fixed number of bits and a universal format for all irises.

6. The method of claim 5, wherein said analysis step includes:

applying said bandpass filters to a region cf the raw iris image signal to remove luminance bias, to remove slow luminance gradients arising from oblique illumination, and to remove noise, and to prevent aliasing.

7. The method of claim 6, wherein:

the value of each bit in the iris code is specified as a "1" or "0", by computing over whatever region of the iris constitutes the support for the said filters of a given size at a given location, the most-significant-bit of the filtered outputs according to the following definitions;

space="preserve" listing-type="equation">MSB.sub.Re (r,θ

)=1 if Re∫

.sub.ρ

∫

.sub.φ

e.sup.-2π

iω

(θ

-φ

) e.sup.-(r-ρ

).spsp.2.sup./α

.spsp.2 e.sup.-(θ

-φ

).spsp.2.sup./β

.spsp.2 I(ρ

,φ

) ρ

dρ

dφ

>

0

space="preserve" listing-type="equation">MSB.sub.Re (r,θ

)=0 if Re∫

.sub.ρ

∫

.sub.φ

e.sup.-2π

iω

(θ

-φ

) e.sup.-(r-ρ

).spsp.2.sup./α

.spsp.2 e.sup.-(θ

-φ

).spsp.2.sup./β

.spsp.2 I(ρ

,φ

) ρ

dρ

dφ

≦

0

space="preserve" listing-type="equation">MSB.sub.Re (r,θ

)=1 if Im∫

.sub.ρ

∫

.sub.φ

e.sup.-2π

iω

(θ

-φ

) e.sup.-(r-ρ

).spsp.2.sup./α

.spsp.2 e.sup.-(θ

-φ

).spsp.2.sup./β

.spsp.2 I(ρ

,φ

) ρ

dρ

dφ

>

0

space="preserve" listing-type="equation">MSB.sub.Re (r,θ

)=0 if Re∫

.sub.ρ

∫

.sub.φ

e.sup.-2π

iω

(θ

-φ

) e.sup.-(r-ρ

).spsp.2.sup./α

.spsp.2 e.sup.-(θ

-φ

).spsp.2.sup./β

.spsp.2 I(ρ

,φ

) ρ

dρ

dφ

≦

0

8. The method of claim 7, wherein said comparing step includes the steps of:

comparing any two iris codes by computing the elementary logical XOR (exclusive-OR logical operation) between all their corresponding bits; and

computing the squared norm of the resulting binary vector;

wherein this comparison measure is defined as the Hamming distance between the two iris code vectors.

9. The method of claim 8, further comprising:

repeating the comparing step for several different relative shifts of the iris code along its angular axis, to compensate for possible tilt of the head of the presenting subject or torsional eye rotation.

10. The method of claim 9, wherein said calculating step includes the steps of:

converting said Hamming distance into a calculated likelihood that the two codes originated from the same iris, and hence the same person.

11. The method of claim 10, wherein:

said calculated likelihood is found by computing the probability that an observed matching fraction of bits in the presenting code and reference code could match by chance if the codes were independent, i.e., arising from separate irises.

12. The method of claim 11, wherein:

the measured Hamming distance is converted into a probability that the two said iris codes are from the same eye;

a preselected criterion is applied to said measured Hamming distance to generate a "yes" or "no" decision; and

a confidence level for the decision is provided by the calculated probability.

13. The method of claim 12, wherein:

said circular pupillary boundary is defined by the relationship;

##EQU7## where r is the radius of said boundary, x₀ and y₀ are center coordinates, and I is image intensity;

wherein said radius and center coordinates are varied in a preselected pattern.

14. The method of claim 13, wherein:

said other circular boundary between the iris and sclera portion of the image is defined by determining the distances from said coordinate system origin to the left and right limbus in accordance with the following relationship;

##EQU8## where r is the pupillary radius previously defined, δ

is a radial shell distance, I(p,θ

) is the original image intensity in polar coordinates, and φ

equals 0 or π

in order to find the limbus in the 3-o'"'"'clock or 9-o'"'"'clock meridia, respectively.

15. The method of claim 14, wherein:

said analysis bands include a plurality of annular bands extending completely around said pupil, and a plurality of semi-annular bands extending from polar angles of approximately 45 degrees to 135 degrees and from 225 degrees to 315 degrees around said pupil.

16. The method of claim 15, wherein:

said analysis bands are spaced at equal fractional radial distances from an inner point located at a preselected fractional distance from said pupillary boundary to an outer point located at preselected fractional distance from said limbus, said spacing being angularly weighted to account for said difference between fractional distances from said coordinate center to said right and left limbi.

17. The method of claim 12, wherein:

said bandpass filters are 2-D Gabor filters.

18. The method of claim 17, wherein:

said 2-D Gabor filters are defined in polar coordinates as follows;

space="preserve" listing-type="equation">G(r,θ

)=e.sup.2π

iω

(θ

-θ

0) e.sup.-(r-r0).sup.2 /α

.sup.2 e.sup.-(θ

-74 )) .sup.2 /β

.sup.2

where r is radius, θ

in angular distance in radians, ω

is frequency, and α and

β

are constants.

19. The method of claim 18, wherein:

said iris code is comprised of 2048 bits.

20. The method of claim 1, wherein:

said circular pupillary boundary is defined by the relationship;

##EQU9## where r is the radius of said boundary, x₀ and y₀ are center coordinates, and I is image intensity;

wherein said radius and center coordinates are systematically varied by iterative gradient ascent to find the maximum in said defined relationship.

21. The method of claim 20, wherein:

said other circular boundary between the iris and sclera portion of the image is defined by determining the distances from said coordinate system origin to the left and right limbus in accordance with the following relationship;

##EQU10## where r₀ is the pupillary radius previously defined, δ

is a radial shell distance, I(p,θ

) is the original image intensity in polar coordinates, and φ

equals 0 or π

in order to find the limbus in the 3-o'"'"'clock or 9-o'"'"'clock meridia, respectively.